Building on the BIRN Workshop BIRN Systems Architecture Overview Philip Papadopoulos – BIRN CC, Systems Architect.
Post on 18-Dec-2015
235 Views
Preview:
Transcript
Building on the BIRN Workshop
BIRN Systems Architecture Overview
Philip Papadopoulos – BIRN CC, Systems Architect
Key Systems Challenges
Large-scale data is distributed on a National Scale• How do you easily locate what you want?• How do you translate it to what your SW tools understand?• Where do you analyze it? • How do you move it efficiently? • How do you secure it to properly limit and log access?
The underlying software systems are complex• How effectively can this complexity be hidden?
Software technology continually evolves and BIRN must adapt
Goal: provide a systems “cookie-cutter” for adding new, secured, resources to form a federation
Meta DataCatalog
PortalServer
SoftwareServer
BIRN CC
A View of BIRN Federated Data
Multi TB Disk array
StorageServer
DB Server
AccessControl
MRI Images
Mouse DB-B
EM Images
Access
Access
Mouse DB-D
Histology
Access
Mouse DB-C
2 Ph. Img
Access
Mouse DB-A
EM Images
BIRN User
? Give me an index of all DAT-KO Striatum Images
Federated data may be in a variety of representations
• databases
• image files
• simulation files
• flat text files
• …
BIRN Data Components
Per Site• Raw, processed and annotated data products• Access policy for locally-held data (authorization)
Distributed• Distributed File System
Location by file name
• Simple Data Federation (BIRN Virtual Datagrid – SRB) Location indexed by content. Collection management
• Integration across databases (Mediator) Translation among data vocabulary
Authentication and Encryption• Grid Security Infrastructure (GSI) as a base
BIRN Core Software Infrastructure
GASS GRAMGSI SRBGridFTPNWSMDS/GRIS
File/DataJob ManagementAuthentication Information
Grid Middleware Services
BIRN Portal
Data SourcesApplicationsComputation
Remote Servers / Sites
Custom APIs mediator client
GridPort Services
registry planner
PortletsGrid Services
Web Server / Applications Server
Collaboration Data ManagementViewing/VisualizationPipelines Queries/Results
BIRN Toolkit
mediatorgatew ay
executor
Statistics/Analysis Spatial Ontology
PACS
Distributed Resources
Distributed Resources
• BIRN builds on evolving community standards for middleware
• Adds significant new capability in allowing disparate data to be queried
• Integrates domain-specific tools to be aware of a much larger data and resource space
• Utilizes commodity hardware and Internet2 for baseline connectivity
Hardware and Software Provisioning
BIRN Coordinating Center• Provide some BIRN-wide services
Portal server, meta data catalog, Public Key Infrastructure PKI certificate server
• All can be replicated for resiliency and performance
• Network and remote-resource monitoring• Software Integration and complete software suite
BIRN testbed resources• A BIRN “Rack”
Access control Data storage Networking monitoring
Existing resources form a solid backbone for extension
Getting Started:BIRN Rack Assembly
Getting Started:BIRN Rack Assembly
Initial rack of equipment was prescribed, but designed from commodity components• Treated as a cluster of specialized configurations
BIRN CC, assembled, tested, shipped, and installed
N2400 NAS
1 - 10 TB
Cisco 4006
DL380 Grid POP
DL380 - Network Stats
GigE Net Probe
APC UPS
DL380
NCRR
Monitoring what is out there
Baseline System Monitoring using ganglia (http://metabirn.nbirn.net)
Gives a live snapshot of standard system metrics• CPU Load
• Network Load
• Disk Utilization
Also need to monitor data utilization (currently 2.6M files, 2.1 Terabytes)
Provisioning a consistent set of Software
Rocks Cluster Toolkit provides and automated baseline configuration for all racks and CC services• Downloadable CD Set to build Linux systems• Configuration is easily transferred to new HW• Uses the NSF Middleware Infrastructure (NMI)
software for common grid services 100’s of clusters (over 40TF aggregate deployed
worldwide) We’ve created a BIRN “roll” CD that integrates
• BIRN domain tools (e.g. 3DSlicer, LONI Pipeline, FreeSurfer)
• Database (Oracle) and SRB Configuration Commodity hardware and our Linux-based SW
Stack allows us to replicate “cookie cutter style”
Tough enough to keep on ticking
Brigham and Women’s Surgical Planning Lab tried to “flood” a BIRN rack. The rack was OK.
top related